JP2012523670A - Dynamically reconfigurable structure for large battery systems - Google Patents

Abstract

  A dynamically reconfigurable battery structure is provided for managing large battery systems. The structure monitors, reconfigures, and controls a large battery system online. The structure is built on a topology-based bypass mechanism that provides a set of rules that change the battery pack configuration, and thereby a semantic bypass mechanism that is reconfigured to recover battery cell connectivity from battery cell failures. More specifically, the semantic bypass mechanism implements a constant voltage holding policy and a dynamic voltage tolerance policy. The former strategy is effective in preventing inevitable voltage drops during battery life, while the latter strategy is effective in supplying different amounts of power that meet a wide range of application requirements.

Description

(Cross-reference of related applications)
This application claims the benefit of US Provisional Patent Application No. 61 / 168,472, filed Apr. 10, 2009. The entire disclosure of the aforementioned application is incorporated herein by reference.

(Field)
The present disclosure relates to battery management, and more particularly, to dynamically reconfigurable structures for large battery systems.

(background)
The demand for electric vehicles equipped with hybrid drive trains has increased rapidly worldwide, mainly due to the recent surge in fuel prices. As oil reserves continue to decline and crude oil prices rise and no viable alternative fuel technology has become apparent, only the demand for electric vehicles with hybrid drivetrains is increasing. According to a recent survey, even in 2008 alone, 36.0% of drivers worldwide wanted to buy a car with a hybrid drive train, and 45.8% were interested in buying a fully electric vehicle. Show. Electric vehicles are all powered by electrical energy from tens of thousands of battery cells. These battery cells are classified and assembled as a set of battery packs. Individual cells within a pack that must be exposed to and operate in a harsh environment may have different operating characteristics due to differences in manufacturing tolerances, uneven temperature conditions across the pack, or uneven aging patterns. Have. These various settings in turn have a decisive influence on the charging and discharging of the battery cells. Within a series chain of battery cells, a low capacity weak battery cell reaches its full charge state well before the remaining battery cells in the chain. On the other hand, a good battery cell may overcharge if a weak cell cannot reach its full charge due to high self-discharge and / or shorted cells. Within the series chain of battery cells, the open cells open other cells in the chain. All these phenomena eventually lead to battery cell failure, which is unavoidable, especially in large battery packs.

  The most commonly used method for managing large battery systems is module-based, where battery cells are classified into their smaller modules, each of which is managed while a group of modules is managed by a global controller. Monitored, controlled and balanced by corresponding local controller. In such a modular battery management system, individual electronic control units (ECUs) receive information such as cell voltage and current, temperature, etc. on the battery packs connected in series via an equalizer connected to each battery cell. Collect and then process the collected information and report to the central ECU responsible for making the local ECU function as needed.

  Individual battery cells can be charged and discharged separately via the switches around them. However, performing separate discharges or specific cell bypasses requires careful management of battery cell placement and battery dynamics while taking into account their characteristics (eg, cell voltage balance and capacity efficiency).

  Battery cell failure is unavoidable, especially for large batteries, and failure rates for multi-cell battery packs are much higher than that of each cell due to inter-cell interactions and dependencies. Unlike battery packs used in portable electronic devices, the electric vehicle environment places many stringent requirements on battery cells and their management.

“Modeling of galvanostatic change and discharge of the lithium / polymer / insertion cell” J. of Power Sources, 140 (6): 1526-1533, 2003.

  There are two major challenges in developing a dynamic reconfiguration structure for a large-scale battery management system. First, if a battery cell failure is detected, the structure must be able to reconfigure battery connectivity online. Healthy battery cells probably must continue to be used in two hierarchical forms of connectivity: battery cells in each pack (cell level) and packs in the entire battery system (pack level). Second, unlike battery powered portable devices, especially for electric vehicles, large battery management systems have multiple output terminals (from battery packs) that supply different voltages for different applications and / or devices. Need. However, physical separation of the battery pack is rarely an option, mainly for cost reasons.

  This section provides background information regarding the present disclosure that is not necessarily prior art.

(Overview)
The reconfigurable battery system is provided with a plurality of battery circuits adjacent to each other. The battery circuit includes an input terminal, an output terminal, a battery cell with a positive terminal and a negative terminal interposed between the input terminal and the output terminal, and a plurality of switches for interconnecting the battery cell with the battery cell in the adjacent circuit. Multiple switches place battery cells in series with battery cells in adjacent battery cells, place battery cells in parallel with battery cells in adjacent battery circuits, or disconnect battery cells from battery cells in adjacent circuits Can be configured. The control unit receives the output reference and controls the switches in each battery circuit to form a circuit arrangement that meets the output reference while substantially maximizing the power supplied by the circuit arrangement, the output reference being the circuit arrangement Define several outputs for and voltage requirements for each output.

  The plurality of switches include an input switch connected between the input terminal and the negative terminal of the battery cell, a parallel switch connected between the output terminal and the positive terminal of the battery cell, between the negative terminal of the battery cell and the negative terminal of the adjacent battery circuit. It includes a connected bypass switch and a series switch connected between the positive terminal of the battery cell and the negative terminal of the adjacent battery circuit. The plurality of switches further include not only an output terminal switch interposed between the output terminal and the output terminal of the adjacent battery circuit, but also an input terminal switch interposed between the input terminal and the input terminal of the adjacent battery circuit.

  Further areas of applicability will become apparent from the description provided herein. The descriptions and specific examples in this summary are intended for purposes of illustration only and are not intended to limit the scope of the invention.

FIG. 2 is a diagram illustrating an exemplary arrangement for a reconfigurable battery system. It is a diagram showing reconfiguration of a battery cell under a constant voltage policy. It is a diagram showing reconfiguration of a battery cell under a dynamic voltage allowance policy. FIG. 6 is a graph showing the difference between a reconfiguration scheme and a legacy scheme for battery life. FIG. FIG. 6 is a graph showing battery life gain achieved by reconfiguration as a function of the number of battery cells in a series chain. It is a graph which shows presentation of the electric power corresponding to the change of a demand voltage. It is a graph which shows the comparison of the dynamic voltage tolerance in a maximum supplyable electric power, and a constant voltage maintenance policy. Fig. 6 is a graph showing dynamic reconfiguration according to voltage demand for different discharge rates. Fig. 6 is a graph showing dynamic reconfiguration according to voltage demand for different discharge rates. Fig. 6 is a graph showing dynamic reconfiguration according to voltage demand for different discharge rates. Fig. 6 is a graph showing dynamic reconfiguration according to voltage demand for different discharge rates.

(Detailed explanation)
The drawings described herein are intended to be illustrative of selected embodiments rather than all possible embodiments, and are not intended to limit the scope of the present disclosure. Corresponding reference characters indicate corresponding parts throughout the several views. A reconfigurable battery cell can be any cell that can convert chemical energy into electrical energy and vice versa. Typically this is achieved by electrochemical oxidation and reduction reactions. These reactions involve generating an electric current by exchanging electrons between the electroactive species in the two electrodes inside the battery cell via a load. Ideally, the total number of current units from the battery cell, ie the coulomb, is always the same throughout its life cycle. However, in reality, battery cell characteristics are not nearly ideal everywhere due to uncertainties in reaction rates and diffusion processes and / or dissolution of active material in battery cells over time. Typical battery cells are nickel metal hydride (NiMH), lithium ion, nickel cadmium (NiCd), lithium ion phosphate, lithium sulfur, lithium titanate, nickel metal hydride, nickel metal hydride, nickel ion, sodium sulfur, Vanadium redox and rechargeable alkali can be included. The architecture described below can be applied not only to other types of rechargeable battery cells but also to these.

  Rechargeable battery cells actually exhibit different characteristics. For example, the battery terminal voltage is not constant during the discharge, and the voltage drops nonlinearly with the discharge rate. The higher the discharge rate, the steeper voltage drop. For this reason, a DC-DC converter can be used to shift and stabilize the supply voltage. Second, the battery capacity varies with the discharge rate, and the higher the discharge rate, the lower the battery capacity. Third, the battery has a limited charge recovery effect at high discharge rates. A high load current in a short time creates a higher concentration gradient between the electroactive species, making it impossible to use unused charges due to delays between reaction and diffusivity. Thus, when the battery rests for a while at a low (or zero) charge rate, the temporarily dropped voltage will rise. Finally, temperature also affects internal resistance and full charge capacity. The lower the temperature, the higher the internal resistance and lower the full charge capacity. On the other hand, self-discharge occurs due to high temperature, and the actual supply capacity is reduced. In addition to these characteristics, some batteries, such as NiCd batteries, are known to have memory effects not found in lithium ion batteries.

  Apart from the temporary changes in battery capacity described above, batteries can lose some capacity due to undesirable side reactions including electrolyte degradation, active material dissolution, and passive film formation, resulting in high internal resistance. Finally, battery cell failure will occur. There are several possible obstacles that make it difficult to predict battery cell behavior. First, the open circuit can be in a fail-safe mode for other battery cells in the series chain that includes the opened battery cell to limit further damage to other battery cells. However, this failure mode may not be useful for applications because all battery cells in the series chain may open and become unusable. Second, shorts with abnormally low electrical resistance are almost voltage so that the remaining battery cells in the chain can be overloaded somewhat while the entire battery pack (ie, a set of battery cells) continues to function. There is no descent. Finally, the possibility of explosion is avoided through a protection circuit that detects and stops extremely high currents.

  FIG. 1 shows a typical arrangement for a reconfigurable battery system 10. The reconfigurable battery system 10 is generally composed of a plurality of battery circuits 30a-30n arranged adjacent to each other. Each battery circuit 30a-30n has an associated control module 20a-20n. In the exemplary embodiment, the control modules 20a-20n are implemented by the controller 50, but it is contemplated that the functions supported by the control modules 20a-20n (or portions thereof) may be divided among multiple controllers. As used herein, the term module refers to an application specific integrated circuit (ASIC), electronic circuit, processor (shared, dedicated, or group) and / or memory that executes one or more software or firmware programs. (Shared, dedicated, or group), combinatorial logic, and / or other suitable components that provide the described functionality, parts thereof, or may be included. It should be understood that the software or firmware program is implemented as computer-executable instructions that reside in computer memory and is executed by a computer processor.

The battery circuits 30a-30n include an input terminal 36a, an output terminal 34a, and a battery cell 32a interposed between the input terminal 36a and the output terminal 34a. The design of the dynamic reconfiguration structure is guided by the principle that any battery cell must be able to be bypassed. In addition, as few switches as possible should be placed around a defined battery cell to minimize cost and improve reliability. In the exemplary embodiment, each battery circuit further includes four switches (also referred to as S _I) input switch 38a that connected between the negative terminal of the input terminal 36a and the battery cells 32a, the output terminal 34a (also known as S _P) connected parallel switches 44a between the positive terminal of the battery cell, (also referred to as S _B) negative terminal and a bypass switch 40a connected between the negative terminal of the adjacent battery circuit of the battery cells, and a battery a positive terminal and the adjacent battery circuit of the negative terminal between the connected series switch 42a of cell 32a (also called S _S). Battery circuits 30a-30n, as described later, another by the input terminal switches 46a-46n and the output terminal switches 48a-48n battery systems to provide a plurality of terminals _(each also referred to as _{S IT} and _{S OT)} Connected. While reference is made to specific switch arrangements, other switch arrangements also fall within the broader aspects of the present disclosure.

  The control units 20a-20n configure the switches in the plurality of battery circuits to form different circuit arrangements. For example, a battery cell can be configured in series by setting the switches in a defined battery circuit 30b as follows: the input switch 38b is set off and the series switch 42b is turned on. , The bypass switch 40b is set to OFF, and the parallel switch 44b is set to OFF, where ON is a closed circuit between both ends of the switch and OFF is an open circuit between both ends of the switch. When a plurality of cells are arranged in series, the cell 32b can be bypassed by setting the switch as follows: the input switch 38b is set to OFF, the series switch 42b is set to OFF, and the bypass Switch 40b is set on and parallel switch 44b is set off. It will be readily appreciated that a switch in the battery circuit at either end of the series string can be configured to place each cell in the series string or to be bypassed.

  In order to configure the battery cells in parallel with each other, the switches in the defined battery circuit 30b are configured as follows: the input switch 38b is set on, the series switch 42b is set off, and the bypass Switch 40b is set off and parallel switch 44b is set on. Similarly, a switch in the battery circuit at either end of the parallel grouping can be configured to place each cell in parallel with the remaining cells. When a plurality of cells are arranged in parallel, the cells can be bypassed by setting all the switches in the defined battery circuit to OFF.

The architecture of the dynamically reconfigurable battery system 10 can be represented as ψ = (E, F, S, D), where E is an array of sensors, {E _{1,. . . ,} E _{i,. . . ,} E _k }, each of which reads the voltage and current of the corresponding battery cell. F denotes an array of feedback switches that maintain a decision on which cells the controller should bypass, {F _{1,. . . , Fi,. . . ,} F _k }. When a battery cell failure is detected in device i, (F _i , On) is turned. S denotes an array of switches, {S _{1,. . . , Si,. . . ,} S _k }, where S _i is composed of S _{i, I} , S _{i, 0} , S _{i, B} , S _{i, S} , S _{i, P} , S _{i, IT} , and S _{i, OT.} . D is a set of battery devices, {D _{1,. . . , Di,. . . ,} D _k }. The connectivity of these devices can be considered as an n _s × n _p matrix.

Here, n _s is the number of battery cells connected in the series chain, and n _p is the number of series chains connected in parallel. The terms V _d and V _a indicate the voltage demand and average voltage of the battery cell (or set of battery packs), respectively. The voltage demand is determined by the application. Similarly, f _N is defined as follows.

Here, (F _i ) is an indication function, that is, (F _i , Off) holds, the function returns to 1, otherwise returns to 0.

During operation, the control unit for the defined battery circuit monitors the operating state of the battery cells in the battery circuit and controls the switches in the battery circuit according to the operating state. In the exemplary embodiment, control units 20a-20n communicate with two sensors 54a-54n and 56a-56n to monitor battery status. For example, the control unit 20a-20n senses the state of charge (SOC) and voltage of its battery cell via sensing devices 54a-54n and 56a-56n. The SOC of the battery cell can be estimated by measuring and integrating the current flowing into and out of the battery cells 32a-32n over time, and is called coulomb count. In practice, voltage and temperature can also be taken into account as battery variables. Thus, the function f _V , _T (SOC, ∫Idt) based on the contents of the coulomb count returns to the SOC. On the other hand, in general, direct voltage measurements are not accurate enough to use as indicators due to their dependence on discharge rate and temperature. In some embodiments, a Kalman filter can be applied to estimate the voltage. Alternatively, it can be assumed that an integrated functional function f _V , _I , _T (SOC, ∫Idt) is given and returns to [V, SOC]. Other techniques for determining the state of charge and / or voltage of a battery cell are within the scope of this disclosure. It will be readily appreciated that different types of sensors can be used to monitor battery status.

At periodic monitoring intervals (Δt), the controller 50 checks the SOC of each battery cell via the corresponding control unit 20a-20n,

Rotation is triggered even when the following holds, where δ is a threshold value that determines the boundary of the maximum SOC variation. The larger δ, the more battery cells become unbalanced. Furthermore, since the variation increases as Δt increases, the variation needs to be adjusted by δ together with Δt. In particular, Δt is inversely proportional to the discharge rate. A rotation event is an adjustment in a battery pack in which a healthy battery cell is rotated along with other healthy battery cells in order to keep the cell healthy.

For purposes of discussion, a failed cell can be considered a battery cell that can only charge up to 80% nominal capacity and / or have a low voltage up to a cutoff voltage in a fully charged state. Thus, when the battery cell i is determined to be faulty, (F _i , On) is turned in the control unit i. Other criteria for determining the failed cell are also conceivable.

At each monitoring interval (Δt), the controller 50 also checks the average voltage,

If does not hold, it triggers a reconfiguration event, where α specifies the upper limit of voltage imbalance. It can be observed that α is adjusted based on the granularity within the supply voltage. A reconfiguration event causes the controller to change the topology of the battery circuit. A reconfiguration event typically occurs when a battery cell is determined to have failed. A reconfiguration event may also occur when an additional application requires a voltage supply and thus requires a multi-terminal configuration. Other types of triggering reconstruction events are also contemplated by this disclosure.

  In the event of a battery cell failure or another triggering reconfiguration event, the semantic bypass mechanism configures battery connectivity. In general, the semantic bypass mechanism implements a policy of supplying a wide range of voltages while following the voltage balance across the parallel groups of the series chain. In an exemplary embodiment, two policies are implemented by the semantic bypass mechanism, although other policies are contemplated by the present disclosure. The semantic bypass mechanism is realized by a controller.

  First, the constant-voltage-keeping policy specifies to keep the supply voltage as constant as possible over the life of the battery despite battery cell failure. For this purpose, the series chain containing the failed battery cells is bypassed. However, the voltage of both healthy use and non-use battery cells in the series chain are separated over time, creating an unbalanced voltage between the battery cells in the series chain. For this reason, a rotation event is triggered during monitoring to reconfigure battery cell connectivity. For connectivity reconfiguration, the lowest level battery cells of those SOCs are selected first.

  FIG. 2 illustrates the reconfiguration of battery cells in a series chain under a constant voltage holding policy. In FIG. 2, two configurations 60a and 60b are shown. In the first configuration 60a, the last four battery cells are connected in series. In the second configuration 60b, the controller 52 connects the four central battery cells in series according to the constant voltage holding policy. As can be seen, the healthy battery cell has been rotated and therefore keeps the voltage constant. As can be observed from the figure, the fault cell 62 is excluded from both configurations.

In order to implement a constant voltage holding strategy, the controller 52 must determine how many battery cells must be bypassed. The number of battery cells to be bypassed is calculated as follows. When Vd is given, first, n _s

And the use of Va cancels the non-linear voltage drop during their lifetime. next,

N _P is derived from where f _N (Ψ) indicates / returns the total number of available battery cells. This equation leads to healthy battery cells to be bypassed (f _N (Ψ) −n _s · n _p ). This procedure is repeated at regular intervals (Δt) or at the start of a reconfiguration event.

Alternatively, a dynamic voltage that supports as many applications as needed and, given available battery cells, improves the maximum available power at the expense of a voltage drop corresponding to a single battery cell voltage. An acceptable policy (dynamic-voltage-allowing policy) is defined. Under a dynamic voltage tolerance policy, one or more healthy battery cells in the series chain can be selected as shown in FIG. To apply this policy, n _p is left fixed according to application requirements, then

N _s is calculated, and the result (f _N (Ψ) −n _s · n _p ) is a healthy battery cell to be bypassed. Similar to the constant voltage holding policy, the battery cells are selected based on their SOC. Similarly, this procedure is repeated at the start of a periodic monitoring interval (Δt) or a reconfiguration event.

These two policies can be further understood from the examples described later. Assuming three parallel groups, each with four series battery cells, the configuration is expressed as:

Here, O indicates which of the corresponding cells is used. Assume that the voltage of each cell is equal to 1V so that each series string outputs 4V. Assuming C6 and C8 fail, the two cells in the other group must rest to balance the voltage across the group, resulting in this configuration.

Here, X indicates that the corresponding cell fails, and − indicates that the cell is resting.

The semantic bypass mechanism reconfigures battery connectivity according to one of two strategies. When the constant voltage holding policy is applied (ie, the demand voltage is 4V), the final configuration is as follows:

The power output by this configuration is 4V ^* 2 (= 8P).
On the other hand, when the dynamic voltage tolerance policy is applied, the final configuration is:

The power output by this configuration is 3V ^* 3 (= 9P). In the dynamic voltage tolerance policy, the number of parallel groups (n _p ) does not change. Instead, the available power can be increased by adjusting the output voltage in ns.

These two policies complement each other to maximize battery usability. In particular, the constant voltage holding policy is applied whenever necessary according to the load demand of a specific application. For example, a system that supports dynamic voltage scaling reduces its system voltage when it has low demand. To do this, systems often use a step-down DC-DC converter. The DC-DC converter may consume energy while the conversion is taking place. In this case, if the constant voltage holding policy is applied to a reconfigurable system instead of using a DC-DC converter, further energy saving is possible. On the other hand, the dynamic voltage holding policy is applied when a plurality of applications should be accommodated simultaneously. Applications that require different voltages and powers require specific capacities and therefore require the definition of the number of parallel groups n _p and they are fixed.

Once n _s and n _p are determined according to the applied policy, controller 50 applies the connectivity configuration algorithm to achieve the desired circuit placement. A typical construction algorithm is described below.

The connectivity configuration algorithm begins by determining which cells in the system can be used. When the kth local control unit reports a cell failure to the controller, the controller permanently bypasses the failed cell by updating its data structure PB and setting PB (k) to 1, where PB is the battery cell's A bit vector having a size equal to the total number, and k is the kth bit of the bit vector. When a cell failure occurs, healthy cells in each parallel group become inactive, i.e., they need to be temporarily bypassed. This intentional bypass is tracked by a controller that maintains another data structure TB for temporary bypass, where TB is a bit vector of a size equal to that of PB.

Starting with the first cell available, a parallel connection group is constructed by the algorithm. Each cell is evaluated in turn. Healthy cells are connected in series, and healthy cells reflected by unhealthy or failed cells or TBs are bypassed. This procedure is repeated until the healthy cells of the n _s are connected in series, parallel connection group is formed. This process moves to the next parallel connection group and is repeated until n _p parallel connection groups are formed. Other processes may be used to connect the cells to achieve the desired circuit arrangement.

Unlike ideal cells, the cell output voltage is not constant during their discharge. That is, ns × Va decreases nonlinearly and deviates from Vd. The deviation can be handled using a DC-DC converter. However, DC-DC converters dissipate energy in the form of heat. The energy dissipation is measured by the conversion efficiency (EFF _DC-DC ) for a DC-DC converter given by EFF _DC-DC = (I _OUT × V _OUT ) / (I _IN × V _IN ), where I _IN and V _IN are Each is the input current and voltage to the DC-DC converter, and I _OUT and V _OUT are the output current and voltage therefrom, respectively. If the input variation is not extreme, the EFF _DC-DC can be approximated between 75-95%, and the DC-DC converter is most efficient when the input voltage is closest to the output voltage. It is important to find an appropriate point in time for reconfiguration to minimize power dissipation. F _DC-DC : V _IN × V _OUT → EFF _DC-DC which defines a function of power. Next, power dissipation is given by PD = (1−f _DC−DC (V _IN , V _OUT )) × V _IN × I _IN . Once _VIN is determined, a constant voltage holding policy is applied.

To minimize power dissipation, to adjust the V _IN to V _d to reconfigure the cell connectivity. Assuming that the reconfiguration overhead is limited by the switching overhead, the switching overhead includes power consumption that signals the switch to turn the switch on and off. This overhead in discrete time rarely fluctuates and is approximated to be a constant amount of power dissipation that is negligible compared to energy dissipation in continuous time. The controller self-configures the cell placement whenever the prior power dissipation is greater than the post-power dissipation (after reconfiguration) including switching overhead, ie, when the following conditions are met:

Here, V _C represents a prior terminal voltage, V ^* d is an approximation of the demand voltage Vd, and α is a switching overhead. With V ^* d, the controller recalculates _ns and _np . This criterion allows the controller to reconfigure cell connectivity in real time.

As described above, the above-described architecture can be extended to a plurality of battery packs configured by the battery system 10 in which each battery pack is reconfigurable. In other words, each battery pack includes a plurality of battery circuits and a local controller that controls the operation of the battery circuits. The extended architecture further includes a global controller that performs data communication with each local controller to coordinate functions between battery packs. The extension of the dynamic reconfiguration structure to multiple battery packs is denoted as Y = (E, F, S, Ψ), where Ψ = {Ψ _{1,. . . ,} Ψ _{2,. . . ,} Ψ _k }. Y is constituted by a global controller cooperating with the local controller, and each Ψ _i is constituted via each local controller. The two policies described above are implemented and applied within the global controller.

The global controller, along with the local controller, reconfigures the battery cells in Y and generates a wide supply voltage for the load. Given V _d , the global controller calculates the number of battery cells Ψ _k and Y connected in series, ie

here,

Is the number of battery cells in the series chain in Ψ _k ,

Is the number of battery cells in the series chain in Y. This equation is

If

Or

If

It is established under the condition. After n _s and N _S are solved for the condition, n _{p in} Ψ _k is

Is calculated by Similarly, N _p in γ (the number of series chains connected in parallel) is

Is calculated by As a result, the local controller and the global controller apply connectivity configuration algorithms with arguments of (Ψ, n _s , n _p ) and (γ, N _S , N _P ), respectively, so that all battery cells inside and outside the battery pack Configured in tandem.

When multiple battery packs are arranged in parallel, it may be necessary to bypass one of the battery packs if the battery pack includes a failed battery cell. Similar to the local controller, the global controller determines ψ _i as a failure when (F _i , On) is detected. However, if Ψ _i is simply bypassed, some battery cells in Ψ _i may become unusable. To address this issue, the global controller implements a pack level bypass decision algorithm. In this algorithm, the global controller finds the minimum number of cells that can be used throughout the pack and shows it as _nm , and then how many cells are bypassed in each pack based on the previous value of _nm calculate. This decision is systematically performed by the global controller via the following decision function.

Two examples are provided to better understand this decision function. In the first example, assume that there are four battery packs, each initially having 6 cells. That is, [6, 6, 6, 6]. If 1, 2, and 2 cells have failed in packs 1, 2, and 3, respectively (denoted [5, 4, 4, 6]), _nm = 4 is obtained. The number of bypass to be cells in each pack (shown as _{n b),} i.e., (1,0,0, and 2) the sum of smaller than _{n m} × 2, the algorithm does not bypass the packs 2 and 3. Instead, it decides to bypass 2 cells in pack 1 and 1 cell in pack 4. In the second example, [4, 2, 3, 6] is assumed. In this example, n _m = 2. Since n _b = 7 (ie, 2 + 0 + 1 + 4) is greater than _nm , pack 2 is bypassed to [4, 3, 6]. In this step, n _m = 3. Since n _b = 4 (ie, 1 + 3) is less than n _m + n _m ^* (previous n _m ), the algorithm returns to n _m (ie, last n _m ). Each pack then bypasses the cell based on _nm . That is, 1, 2, 0, and 3 cells in packs 1, 2, 3, and 4 are bypassed, respectively. In this way, the global controller can determine when and how to bypass the battery pack. When the local controller in each pack receives the latest value of _nm from the global controller, each local controller applies a constant voltage holding policy based on _nm .

The reconfigurable structure described above can also be used to support multiple applications where one application requires power from the battery system. For example, in a vehicle, starter motors, windshield wipers and radios may all require power from a battery system. Each application may be assigned a priority defines the output voltage requirements V _d. The output voltage requirement V _d for application k determines the number of cells in series N _{s, k} necessary to satisfy the requirement. N _{p, 1,} N _{p, 2,. . . ,} N _{p, k,} gives the total number of healthy battery cells. N (γ) is a q parallel group for all applications.

It can then be assigned to each application that requires a battery cell. When the number of battery cells necessary to satisfy the application requirement condition exceeds the number of usable battery cells, usable battery cells are assigned to the application based on the priority assigned to the application. If the number of available battery cells exceeds the number of battery cells required to meet the application requirements, the controller can allocate the remaining cells. In either case, the available battery cells are initially distributed to high priority applications, ie, high demand voltage applications. This distribution continues until the remaining cells are not sufficient to be distributed. A typical allocation policy is defined as follows:
Multiple terminal base grouping

In this way, the controller allocates N _{s, k} × N _{p, k} power supplies to each application k. In the case of an extended reconfigurable structure, the allocation policy is implemented by the global controller.

When a battery cell is assigned, the controller configures the battery system accordingly. The battery system is configured to provide input and output terminals for each application having initially assigned battery cells. To do so, input terminal switches 46a-46n and output terminal switches 48a-48n are controlled to provide a plurality of terminals. For example, when the input terminal switch ( _{Si, I,} On) and the output terminal switch ( _{Si, P,} On) are stopped for all battery circuits 30a-30n, the interface to the battery pack is single input terminal and single It has an output terminal. Conversely, the input terminal switch and output terminal switch selection can be set off to segment the battery circuits 30a-30n to provide multiple terminals. For each application segment, the battery system is then configured to meet application requirements using the semantic bypass mechanism described above.

For example, large battery cells for EV and HEV have n _s battery cells connected in series to provide the required supply voltage and n _p parallel groups connected in parallel to determine current (I) flow and It is stuffed to become capacity. Due to the non-linearity of the battery, it is not possible to simply derive the capacity with the ideal battery capacity equation,

Here, T is a discharge time (battery life). Instead, the empirical Poicart relationship models non-linearity in the case of constant current by introducing empirical parameters such as C≡T · I ^α , where α> 1 is called the Poicart value Typically in the range between 1.2 and 1.4.

For the purpose of a reconfigurable battery management system, non-linearities can be modeled using current flow discretization. That is, a real-world system is characterized by a load that varies over time. Such a variable load can be approximated by a piecewise constant load represented by a set of M current levels (i _1,..., I _M ), where M is used to characterize the load, It is determined by the quantization interval Δt (≡t _i −t _{i + 1} ), which is a fraction of the operation time T. That is,

Where 1 _A (t) is a defining function. Therefore, the accuracy in load characterization increases as Δt decreases. When Δt≡T, the load is constant. The load pattern is obtained through empirical measurements and is the discharge profile for the battery cell or pack of battery cells. Thus, the model of equation (6) is generalized as follows:

The total load is the sum of the currents loaded from the individual parallel groups, i.e. I = I ₁ +. . . + I _i +. . . + I _np , evenly distributed at some point in time within the specified tolerance threshold for mismatch, I = n _p · I _i . As a result, the following equation is obtained.

When a cell failure occurs, the number of usable parallel groups is equal to n _p −N (t), where N (t) is the total number of failures that occurred in the battery cell array by time t. In the ABS, the number of usable parallel groups is defined as follows.

Since the number of these failures that occur at discrete time intervals is independent, N (t) is a Poisson distribution with the battery cell failure rate λ distributed. Therefore, the average total number of cell faults occurring up to time t is proportional to t and becomes λ · t. This equation is applied to equation (9) to obtain

here,

It is.

On the other hand, in the legacy scheme, the load on the series chain of operational battery cells increases in proportion to the total number of cell faults in the entire _np parallel group as follows:

The linear increase in load is due to the fact that it fails to reuse any healthy battery cell in the series chain that contains the failed cell. Therefore, the available capacity according to the legacy scheme is calculated by

here,

It is. Therefore, the life gain on legacy schemes as λ is high increases, it is also inversely proportional to the series of battery cells number n _s.

One of the two policies described above is applied based on the configuration of battery cell connectivity. In order to maximize battery cell usage, the capacity of power supplied by all battery cells is selected as a criterion for comparing the two strategies. If m × n matrix is n _p parallel groups represents a combination of n _s battery cell in the serial chain, any element of the battery cell in the matrix is assumed to be defective independently of the others. For example, if one battery cell fails, ( _ns -1) · n _{p of power} is provided based on a dynamic voltage tolerance policy. For simplicity, it is assumed that each battery cell element is capable of 1V and 1A during power _ns · (n _p −1) based on a constant voltage holding policy. Therefore, the break-even point in the policy selection is found when n _s = · n _p . When two or more battery cells fail, the number of battery cells that are left unused due to failed cell bypass reflects the capacity criteria. In other words, the ratio (r) of the number of columns (c) and the number of rows (w) on the failed cells in the matrix can be a factor in making a judgment comparable to the total size of the matrix. Therefore, the breakeven point is determined by the following formula,

here,

If so, a dynamic voltage allowance policy is selected, which provides more power capacity than a constant voltage hold policy.

  The evaluation method is first described, and then the performance of the described architecture is evaluated in comparison to legacy schemes where battery cell connectivity cannot be configured online. Metrics used to evaluate battery performance include battery life and supply voltage. The lifetime is proportional to the total capacity of the battery cell / pack, and the supply voltage determines the power that can be supplied. Battery dynamics were simulated using Dualfoil, which is widely used to design multiple battery systems. See Non-Patent Document 1 for a more detailed description of Dualfoils. Using Dualfoils is sufficient to show how battery connectivity is dynamically reconfigured.

While the reconfiguration structure effectively masks the effects of battery cell failure and extends battery life, the legacy scheme suffers significant battery capacity loss, thus reducing life. Battery life is calculated by the maximum available power and the amount of current continuously drawn from the battery pack. Obviously, the more battery cells fail, the higher the reduction in battery life. FIG. 4 shows a comparison result of battery life. Clearly, legacy schemes lose a significant amount of life as the number of failed battery cells increases. The reason is that the failure of one battery cell causes the series chain containing the failed battery cell to be lost. In contrast, the reconfiguration structure reuses the remaining healthy battery cells in the series chain as backup cells. Thus, if there are more battery cell failures in the other chains, they are replaced with surviving healthy battery cells. FIG. 4 shows the fault tolerance function of the proposed reconstruction structure. For example, if λ · t≡6 to 9 and λ · t≡12 to 15, the life of the battery pack remains constant regardless of the increase in the number of battery cell failures. The difference in lifetime between the two mechanisms increases as the battery cell failure frequency increases. As can be seen from FIG. 5, the lifetime gain achieved by the reconfiguration structure increases substantially with the increase in the number of battery cells (n _s ) in the series chain in each parallel group, increasing the availability of the backup battery cells. This is effective even when two battery cells are connected in series (ie, n _s = 2), and a gain factor of 5 is achieved. Clearly, the more battery cells in series, the greater the gain.

  The dynamic voltage tolerance policy aims to meet a wide range of supply voltage demands from different applications while maximizing the power available. FIG. 6 shows the corresponding maximum resulting from a 25 battery cell pack based on a configuration that sets the demand voltage and the actual supply voltage and capacity of each battery cell to 3.6 V and 1.3 AH, respectively, with an acceptable jitter of 2.5%. It shows the change in suppliable power. Maximum available power is limited between estimated 114W and 120W. This power is a combination of 5 parallel groups and 5 battery cells in the series chain within each group (ie,..., (5, 5)), or one parallel group with 25 battery cells in series (ie 25, 1). Interestingly, a good range of supply voltage, corresponding to the circled group in FIG. 6, is provided while keeping the maximum available power appropriately constant. This means that appropriately changing battery connections can improve battery cell utilization while meeting the demands of the underlying application. On the other hand, the connectivity of (9, 2) or (13, 1) appears to be inefficient with respect to battery cell utilization. However, any battery cell failure or voltage drop can be resolved by substantially replacing them with backup battery cells and maintaining the required voltage level.

Dynamic voltage tolerance and constant voltage holding policies are devised for different purposes, the former aiming to meet a wide range of supply voltage demands, the latter supplying for battery failures or possible voltage drops during battery life The aim is to maintain a voltage tolerance, and together the power available is maintained at a maximum. As such, the two strategies can be compared in terms of available power. FIG. 7 shows the distribution of power magnitude between constant voltage holding and dynamic voltage tolerance policies. In battery connectivity, if n _s > n _p , the dynamic voltage tolerance policy is effective in supplying the maximum power supply, and if n _p > n _s , the constant voltage holding policy is a better option. Become. The reason for this is that unused battery cells / packs are utilized. Obviously, the breakeven point occurs when n _s = n _p .

As described above, since a voltage drop is unavoidable, a constant voltage holding policy is applied to maintain the supply voltage above the demand voltage while the supply voltage is monitored. The monitoring interval (Δt) is directly related to the degree to which the system suffers to drop below the demand. The higher the frequency of monitoring, the shorter the time an application takes, but the higher the monitoring overhead. FIG. 8A shows the change in supply voltage at two different discharge rates during the lifetime of the 700 battery cell pack. Assume that each battery cell is discharged independently following a simulated lithium ion battery discharge distribution in a configuration that provides an output voltage of 4.3 V and a nominal capacity of 1.3 AH. Assume that the demand voltage (V _d ) for the application is 600V. In FIG. 8B, when the battery pack is discharged at the C rate, when the battery pack is monitored every Δt (= 10), it is detected when the supply voltage drops below V _d at 10 ^−th time intervals, The battery pack connectivity is reconfigured into four parallel groups with 143 battery cells in a series chain, ie (143, 4), providing an estimated 604V. For the C2 rate, in FIG. 8C, the underlying application suffers 5 times more battery capacity loss than at the normal discharge rate. In particular, if the supply voltage drop is steeper, the difference between supply and demand voltage will increase. This case can be improved by reducing the monitoring interval (Δt = 10). As shown in FIG. 8D, halving the monitoring interval (Δt = 5) improves voltage drop on-time detection by 67%.

  The foregoing description of the embodiments is provided for purposes of illustration and description. It is not exhaustive and does not limit the invention. The individual elements or functions of a particular embodiment are generally not limited to that particular embodiment and, where applicable, in compatible and selected embodiments, even if not specifically illustrated or described. Can be used. The same can be changed in many ways. Such changes are not considered to depart from the invention and such modifications are intended to be within the scope of the invention.

  Rather, these embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the scope of the invention to those skilled in the art. In order to provide a thorough understanding of embodiments of the present disclosure, numerous specific details are set forth such as examples of specific components, devices, and methods. It will be apparent to one skilled in the art that the specific details need not be employed, and the examples can be embodied in many different forms, neither of which should be construed to limit the scope of the disclosure. . In some embodiments, known processes, known device structures, and known techniques have not been described in detail.

  The techniques used herein are for the purpose of describing particular embodiments only and are not intended to be limiting. As used herein, the singular forms “a”, “an”, and “the” include plural referents unless the context clearly dictates otherwise. The terms “comprises”, “comprising”, “inclusioning”, and “having” are generic and specify the presence of the stated feature, integrals, steps, actions, elements, and / or components. Does not exclude the presence or addition of one or more other features, completeness, steps, actions, elements, components and / or groups thereof. The method steps, processes, and operations described herein should not necessarily be construed as requiring their performance in the specific order discussed or exemplified unless specifically stated as a performance order. It is also understood that additional or alternative steps can be employed.

Claims (22)

A reconfigurable battery system,
A plurality of battery circuits interconnected with each other, each circuit comprising:
An input terminal;
An output terminal;
A battery cell with a positive terminal and a negative terminal interposed between the input terminal and the output terminal;
An input switch connected between the input terminal and the negative terminal of the battery cell;
A parallel switch connected between the output terminal and the positive terminal of the battery cell;
A bypass switch connected between the negative terminal of the battery cell and the negative terminal of the adjacent battery circuit;
A series switch connected between the positive terminal of the battery cell and the negative terminal of the adjacent battery circuit; and
A reconfigurable battery system comprising: a control unit that monitors an operating state of the battery cell in a plurality of battery circuits and controls a switch in the plurality of battery circuits according to the operating state.   The reconfigurable battery system according to claim 1, wherein the control unit determines a number of usable battery cells from the battery cells in the plurality of battery circuits.   3. The reconfigurable battery system according to claim 2, wherein the control unit receives a voltage output request condition for the battery system, and sets a voltage in the switch in the plurality of battery circuits to satisfy the voltage output request condition. A reconfigurable battery system configured to form an output circuit arrangement.   4. The reconfigurable battery system according to claim 3, wherein the control unit determines that a number of battery cells arranged in series need to satisfy the voltage output requirements and are arranged in parallel with each other. Configuring the switches in the plurality of battery circuits by determining a number of possible battery cell groups from the number of available battery cells, wherein each battery cell group satisfies the voltage output requirements A reconfigurable battery system comprising a series of battery cells arranged in series.   3. A reconfigurable battery system according to claim 2, wherein the control unit receives several parallel cell groups and minimizes the several available battery cells that are bypassed in a circuit arrangement. A reconfigurable battery system that configures the switches in the plurality of battery circuits to form the circuit arrangement to be suppressed.   3. The reconfigurable battery system according to claim 2, wherein the control unit divides the number of usable battery cells by the number of parallel cell groups to form each parallel cell grouping. A reconfigurable battery system in which such cells are determined and the cells constituting the parallel cell grouping are arranged in series with each other. A reconfigurable battery system,
A plurality of battery circuits adjacent to each other, each battery circuit comprising:
An input terminal;
An output terminal;
A battery cell with a positive terminal and a negative terminal interposed between the input terminal and the output terminal;
A plurality of switches interconnecting the battery cells with battery cells in an adjacent circuit, wherein the battery cells are arranged in series with the battery cells in the adjacent battery cells; A plurality of switches arranged in parallel with the battery cell or configured to disconnect the battery cell from the battery cell in the adjacent circuit; and
A control unit that receives the output reference and controls the switches in each of the battery circuits to form the circuit arrangement that satisfies the output reference while substantially maximizing the power supplied by the circuit arrangement; A reconfigurable battery system wherein the output reference includes a control unit that defines several outputs for the circuit arrangement and a voltage requirement for each output.   8. The reconfigurable battery system according to claim 7, wherein the control unit determines a number of usable battery cells from the battery cells in the plurality of battery circuits and determines the usable battery cells. A reconfigurable battery system assigned to each output defined by the output criteria.   9. The reconfigurable battery system according to claim 8, wherein the control unit receives the voltage output request condition for a predetermined output, and outputs a voltage satisfying the voltage output request condition for the predetermined output. A reconfigurable battery system that configures the switches in the plurality of battery circuits to form a circuit arrangement.   10. The reconfigurable battery system of claim 9, wherein the control unit for determining a number of battery cells to be assigned to the defined output includes a number of battery cells arranged in series with the voltage output requirement. And determine a number of battery cell groups that can be arranged in parallel with each other from the number of battery cells assigned to the defined output, each battery cell group having the voltage A reconfigurable battery system comprising a series of battery cells arranged in series that satisfy output requirements.   11. The reconfigurable battery system according to claim 10, wherein the control unit is configured when the number of battery cell groups exceeds the number of battery cells arranged in series satisfying the voltage output requirement. A reconfigurable battery system comprising the switches in the plurality of battery circuits so as to form a circuit arrangement for outputting a voltage that satisfies a voltage output requirement condition.   10. A reconfigurable battery system according to claim 9, wherein the control unit receives a number of parallel cell groups for a defined output and the number of usable batteries bypassed in a circuit arrangement. A reconfigurable battery system that configures the switches in the plurality of battery circuits to form the circuit arrangement that minimizes cells.   13. A reconfigurable battery system according to claim 12, wherein the control unit divides the number of battery cells assigned to the defined output by the number of parallel cell groups to each parallel cell. A reconfigurable battery system in which several cells constituting a grouping are determined and the cells constituting a parallel cell grouping are arranged in series with each other.   14. A reconfigurable battery system according to claim 13, wherein the control unit is enabled for use when the number of cells comprising each parallel cell grouping exceeds the number of parallel cell groupings. A reconfigurable battery system that configures the switches in the plurality of battery circuits to form a circuit arrangement that minimizes the number of battery cells.   8. The reconfigurable battery system according to claim 7, wherein the plurality of switches are further defined as an input switch connected between the input terminal and the negative terminal of the battery cell, the output terminal and the A parallel switch is connected between the positive terminals of the battery cell, a bypass switch is connected between the negative terminal of the battery cell and the negative terminal of the adjacent battery circuit, and the positive terminal of the battery cell and the adjacent battery circuit A reconfigurable battery system with a series switch connected between the negative terminals. A reconfigurable battery system,
A plurality of battery circuits adjacent to each other, each battery circuit comprising:
An input terminal;
An output terminal;
A battery cell with a positive terminal and a negative terminal interposed between the input terminal and the output terminal;
Interconnecting the battery cell with a battery cell in an adjacent circuit, placing the battery cell in series with the battery cell in the adjacent battery cell, and paralleling the battery cell with the battery cell in the adjacent battery circuit; A plurality of switches that can be arranged or configured to disconnect the battery cell from the battery cell in the adjacent circuit, the plurality of switches between the input terminal and the input terminal of the adjacent battery circuit Including an input terminal switch interposed between the output terminal and the output terminal of the adjacent battery circuit,
A reconfigurable battery system including a control unit for controlling the plurality of switches in each of the battery control circuits.   17. The reconfigurable battery system according to claim 16, wherein the plurality of switches include an input switch connected between the input terminal and the negative terminal of the battery cell, and the output terminal and the battery cell. A parallel switch connected between the positive terminals, a bypass switch connected between the negative terminal of the battery cell and the negative terminal of the adjacent battery circuit, the positive terminal of the battery cell and the negative terminal of the adjacent battery circuit A reconfigurable battery system including a series switch connected therebetween.   17. A reconfigurable battery system according to claim 16, wherein the control unit determines a number of usable battery cells from the battery cells in the plurality of battery circuits.   19. The reconfigurable battery system according to claim 18, wherein the control unit receives a voltage output requirement condition for the battery system and forms a circuit arrangement for outputting a voltage satisfying the voltage output requirement condition. A reconfigurable battery system comprising the switch in a plurality of battery circuits.   20. The method of claim 19, wherein the control unit determines that several battery cells arranged in series need to meet the voltage output requirement and from the available battery cells to each other. The switches in the plurality of battery circuits are configured by determining several battery cell groups that can be arranged in parallel, wherein each battery cell group is a series arranged in series that satisfies the voltage output requirements. A reconfigurable battery system composed of battery cells.   21. The reconfigurable battery system of claim 20, wherein the control unit receives a number of parallel cell groups and minimizes the number of available battery cells that are bypassed in a circuit arrangement. A reconfigurable battery system that configures the switches in the plurality of battery circuits to form the circuit arrangement to be suppressed.   22. A reconfigurable battery system according to claim 21, wherein the control unit divides the number of available battery cells by the number of parallel cell groups to form each parallel cell grouping. A reconfigurable battery system in which such cells are determined and the cells constituting the parallel cell grouping are arranged in series with each other.